ON THE SENSITIVE STATE OF VACUUM DISCHARGES. 
605 
so large an extent as to reach the piece of glass, the shadow must have taken a 
direction much more oblique than was actually the case; and furthermore, the extreme 
obliquity of the streams of molecules that caused the shadow renders it very improbable 
that they should all have started normally. 
In this particular instance there was a peculiarity which rendered the experiment 
very interesting, as showing how completely the relief phosphorescence is independent 
of that which comes from the negative terminal. The piece of glass was between the 
tinfoil and the negative terminal, so that the shadow pointed towards the negative 
terminal inst ead of away from it. 
If a very small patch of tinfoil, connected to earth, be placed upon a tube of con¬ 
siderable diameter and not too high exhaust to permit the details of the relief phos¬ 
phorescence to be readily distinguished, it will be found that it produces a bright 
central patch of phosphorescence at the point of the tube exactly opposite to it 
(Plate 27, fig. 13), and that this bright central patch is surrounded by an annulus of 
feebler intensity but considerable breadth ending in a fine bright line of phosphorescence 
serving as its outer edge. The breadth of the whole of this phosphorescent area is 
such that it subtends a very considerable finite angle at the patch of tinfoil, the semi¬ 
vertical angle of the cone of rays being from 20° to 30°. Now in this case the patch 
of tinfoil is so small that it may in considering the direction of the resulting molecular 
streams be taken to be a point, and thus we see that the molecular streams from a 
small elemental area would, if unaffected by any other circumstances than those 
necessarily present in a tube, pass off in all directions comprised within a solid angle 
of finite size (depending probably upon the degree of the exhaust and the violence of 
the discharge), surrounding the normal in an approximately symmetrical way, i.e., 
forming a right cone of which it is the axis.* We are not able to speak definitely as 
to the intensity of the streams in the different directions. Those that proceed strictly 
normally are probably the most intense either from the greater density of the streams 
or the greater velocity of the particles, for we find that there is a very bright patch in 
the centre. But this may be partly due, as we shall see, to the fact that the patch of 
tinfoil has a finite though small area. A more difficult matter to account for is the 
apparently sharp limit which bounds the phosphorescent area on its outer side. It is 
difficult to imagine that there can be an abrupt limit to the angular extent of these 
molecular discharges. The most probable hypothesis is that it is due to a wholly 
* It will probably be objected (and with perfect justice) that we are reasoning as tbougb all the 
molecular streams that leave the gas beneath the tinfoil arrive at the opposite side of the tube and make 
themselves visible there. This is of course not the case, and if we were to take a screen situated very 
much nearer to the tinfoil than is the opposite side of the tube we should no doubt get a wider limit to 
the directions of the molecular streams. But the general reasoning is not affected by this, although it is 
most important to bear in mind that our tests do not exclude the possibility of feebler streams of molecules 
issuing at still greater inclinations to the normal, and that it is only the streams that have a certain 
intensity that are rendered visible by phosphorescence. 
MDCCCLXXX. 4 I 
